Pillar 4 – Monitoring and Evaluation: Ecological Monitoring

There are two types of monitoring that can be used in demonstration reaches, together they allow for a good understanding on both the combined impacts of monitoring interventions and the contributions of individual actions.

Condition monitoring

Trend or condition monitoring is used to report on broad scale patterns of river health through time. In demonstration reaches it focuses on the whole demonstration reach and how the condition of the fish community has changed over time. It will identify general trends in the reaches response to the interventions undertaken but gives no indication of the underlying mechanisms and contributions of the different interventions.

Intervention monitoring

Intervention monitoring focuses on the outcomes of particular on-ground actions, and their contribution to fish responses in the reach. The level of monitoring needed depends on the scale of the intervention being undertaken.

The development of monitoring and evaluation activities will be governed by the rehabilitation goals outlined in the Whole of Life Plan, and the interventions planned. To test these goals and interventions it is important to develop stressor/response conceptual models and hypotheses. These will bring together existing information on how the ecosystem may be expected to function and where the gaps in knowledge might be. Conceptual models can be quite simple diagrams. They are starting points gathering and developing knowledge about how the ecosystem in question may respond to the interventions and are qualitative in nature. However, they allow both the development of hypotheses to be tested and suitable indicators to be measured.

(See examples of conceptual models in Boys et al. 2009, Appendix 5, page 53 “Developing Conceptual Models”).

To suitably test conceptual models and hypotheses it is important to have a clear indication of what environmental indicators will be used to measure the condition and response (physical, chemical or biological). Indicators should be developed according to SMART criteria (Specific, Measurable, Achievable, Realistic and Time- bound). They should be measurable using existing and accepted methods (e.g. electrofishing, trapping etc.), and non-destructive to the ecosystem. Sampling of these indicators will allow for the statistical testing of the hypotheses.

(see Example 5d – Examples of indicators and sampling methods used by existing demonstration reaches).

The choice of experimental design is crucial to the success of the monitoring program. Regardless of the resources available, a poorly designed monitoring program will not provide useful information and can result in a waste of resources. The design has to be tailored to the specific project, but the Boys et al. (2009) document provides very good broad guidance. Decisions about statistical analysis methods should be made before data collection starts and both the sampling design and proposed analysis should be developed with the assistance of a competent biometrician.

The experimental design for demonstration reaches needs to incorporate monitoring over different space and time scales.

Spatial

Different spatial scales will be used to evaluate demonstration reach outcomes depending on the type of monitoring. Condition Monitoring will be carried out with the whole demonstration reach as the management unit. Within the demonstration reach, sites will have to be sub- sampled at the sub-reach scale to account for spatial variability. The number and location of these sub-samples will depend on the particular indicator, its patchiness and the statistical power required.

Intervention Monitoring will require a different spatial scale depending on whether the response is expected to occur at specific sites (e.g. placement of lunkers or large woody debris) or throughout the reach (e.g. cold water pollution or Carp management). It is important to remember, particularly in short demonstration reaches, that there can be confounding effects of multiple interventions on the impact of single interventions if interventions are undertaken too close to each other.

Temporal

Monitoring and evaluation programs associated with demonstration reaches need to be implemented over a number of years if an ecological response to interventions is to be detected. It may take 5 to 15 years to detect changes related to some on-ground activities particularly when considering natural “background” variability associated with extreme events such as floods and droughts. Depending on the indicator in question, there may be some short-term responses to interventions. For example, the placement of large woody debris in a reach may result in a relatively rapid increase in numbers of some species. This may simply be due to the woody debris acting as attraction devices, and actual increases in sustainable population numbers may take considerably longer to detect.

Likely response times will impact on the sampling regime and this needs to be made apparent to stakeholders so that there are no unrealistic expectations.

Overall Design

Although the overall monitoring designs for different demonstration reaches will vary, it is important that they include a period of before or pre-intervention monitoring as well as a period of after or post-intervention monitoring. The pre-intervention monitoring should cover a number of years if possible, to account for natural variability. Post-intervention monitoring may require sampling for a long period of time (e.g. 10 years) depending on the expected time lag for a fish population response to be detected. There should also be a control reach for comparison to ensure that the responses are due to the interventions and not natural variability. Together this will form a Before-After-Control-impact (BACI) design. (see Boys et al. 2009).

In most instances, monitoring programs have been undertaken by jurisdictional natural resource management agencies (usually Fisheries). These agencies have the necessary equipment (electrofishers, nets etc.), the permits and the expertise to use the equipment effectively.

The number of sampling sites varies from reach to reach. For example, the Ovens River Demonstration Reach (100km in length) has 10 sites within the reach (five treatment and five non-treatment sites) together with four control sites in the King River. Boat mounted electrofishers and baited traps are used to sample the fish community.

Hollands Creek Demonstration Reach (20km in length) has seven sampling sites within the reach and four control sites on Ryans Creek. Sampling is undertaken with backpack electrofishers (due to the shallow nature of the creek) and fyke nets. Sampling times vary from reach to reach but is usually undertaken annually in summer.

It is important that regular monitoring reports are produced (e.g. annually) and that these are made available to the steering committee and summaries of the results to the wider community. Both successes and failures should be reported so that an adaptive management approach can be taken. Given the expected time lag in native fish response to most management interventions it may be appropriate to celebrate interim milestones with the community such as length of river re-snagged etc.

Given the long time frame for the monitoring program (10 years or more) it is important to maintain its integrity through a quality assurance program and to ensure ongoing data analysis and data storage.